A low-cost and portable active noise control unit
Noise pollution these days has been known as one of the major problems in our daily lives. As Singapore continues building more housing and buildings, the number of construction sites increases, contributing to noise pollution. These noises affect us mentally and can lead to serious health problems....
| 主要作者: | |
|---|---|
| 其他作者: | |
| 格式: | Final Year Project (FYP) |
| 語言: | English |
| 出版: |
Nanyang Technological University
2021
|
| 主題: | |
| 在線閱讀: | https://hdl.handle.net/10356/149660 |
| _version_ | 1826110523872116736 |
|---|---|
| author | See. Emily Hui Hua |
| author2 | Gan Woon Seng |
| author_facet | Gan Woon Seng See. Emily Hui Hua |
| author_sort | See. Emily Hui Hua |
| collection | NTU |
| description | Noise pollution these days has been known as one of the major problems in our daily lives. As Singapore continues building more housing and buildings, the number of construction sites increases, contributing to noise pollution. These noises affect us mentally and can lead to serious health problems. By borrowing the same technology used in noise-cancelling headphones, Active Noise Control (ANC) can be used to mitigate these noises. The aim of this study is to investigate the different feature extraction techniques and classification methods to improve the performance of the ANC system, reducing as much noise that can be heard outside the construction site as fast as possible. The general idea is analysing noise signal features and pre-training the noise signal with certain features to get the corresponding optimal control filters. For practical application, the senses noise is classified online, and the corresponding optimal control filter is assigned to control the signal. Thus, the accuracy will give an indicator on how effective these methods are. Noise samples are collected and divided into two sets, one for training, one for validation. For the training set, feature extraction and classification are conducted. Validation is conducted on the validation set. The extraction of MFCC along with the MLP and CNN models have shown relatively high accuracy scores. Achieving 95% training accuracy and 89% testing accuracy for MLP, and 98% training accuracy and 91% testing accuracy for CNN. |
| first_indexed | 2024-10-01T02:35:50Z |
| format | Final Year Project (FYP) |
| id | ntu-10356/149660 |
| institution | Nanyang Technological University |
| language | English |
| last_indexed | 2024-10-01T02:35:50Z |
| publishDate | 2021 |
| publisher | Nanyang Technological University |
| record_format | dspace |
| spelling | ntu-10356/1496602023-07-07T17:14:03Z A low-cost and portable active noise control unit See. Emily Hui Hua Gan Woon Seng School of Electrical and Electronic Engineering EWSGAN@ntu.edu.sg Engineering::Electrical and electronic engineering Noise pollution these days has been known as one of the major problems in our daily lives. As Singapore continues building more housing and buildings, the number of construction sites increases, contributing to noise pollution. These noises affect us mentally and can lead to serious health problems. By borrowing the same technology used in noise-cancelling headphones, Active Noise Control (ANC) can be used to mitigate these noises. The aim of this study is to investigate the different feature extraction techniques and classification methods to improve the performance of the ANC system, reducing as much noise that can be heard outside the construction site as fast as possible. The general idea is analysing noise signal features and pre-training the noise signal with certain features to get the corresponding optimal control filters. For practical application, the senses noise is classified online, and the corresponding optimal control filter is assigned to control the signal. Thus, the accuracy will give an indicator on how effective these methods are. Noise samples are collected and divided into two sets, one for training, one for validation. For the training set, feature extraction and classification are conducted. Validation is conducted on the validation set. The extraction of MFCC along with the MLP and CNN models have shown relatively high accuracy scores. Achieving 95% training accuracy and 89% testing accuracy for MLP, and 98% training accuracy and 91% testing accuracy for CNN. Bachelor of Engineering (Information Engineering and Media) 2021-06-06T14:35:10Z 2021-06-06T14:35:10Z 2021 Final Year Project (FYP) See. Emily Hui Hua (2021). A low-cost and portable active noise control unit. Final Year Project (FYP), Nanyang Technological University, Singapore. https://hdl.handle.net/10356/149660 https://hdl.handle.net/10356/149660 en A3086-201 application/pdf Nanyang Technological University |
| spellingShingle | Engineering::Electrical and electronic engineering See. Emily Hui Hua A low-cost and portable active noise control unit |
| title | A low-cost and portable active noise control unit |
| title_full | A low-cost and portable active noise control unit |
| title_fullStr | A low-cost and portable active noise control unit |
| title_full_unstemmed | A low-cost and portable active noise control unit |
| title_short | A low-cost and portable active noise control unit |
| title_sort | low cost and portable active noise control unit |
| topic | Engineering::Electrical and electronic engineering |
| url | https://hdl.handle.net/10356/149660 |
| work_keys_str_mv | AT seeemilyhuihua alowcostandportableactivenoisecontrolunit AT seeemilyhuihua lowcostandportableactivenoisecontrolunit |